diff --git a/altamisa/isatab/parse_assay_study.py b/altamisa/isatab/parse_assay_study.py
index 61414b7..5b8af47 100644
--- a/altamisa/isatab/parse_assay_study.py
+++ b/altamisa/isatab/parse_assay_study.py
@@ -817,7 +817,7 @@ def from_stream(cls, study_id: str, input_file: TextIO, filename: Optional[str]
     def __init__(self, study_id: str, input_file: TextIO, filename: Optional[str]):
         self.study_id = study_id
         self.input_file = input_file
-        self.filename = filename or getattr(input_file, "name", "<no file>")
+        self._filename = filename or getattr(input_file, "name", "<no file>")
         self.unique_rows = set()
         self.duplicate_rows = []
         self._reader = csv.reader(input_file, delimiter="\t", quotechar='"')
@@ -849,7 +849,7 @@ def _read_next_line(self):
         except StopIteration:
             self._line = None
         except UnicodeDecodeError as e:  # pragma: no cover
-            msg = f"Invalid encoding of study file '{self.input_file.name}' (use Unicode/UTF-8)."
+            msg = f"Invalid encoding of study file '{self._filename}' (use Unicode/UTF-8)."
             raise ParseIsatabException(msg) from e
         return prev_line
 
@@ -859,7 +859,7 @@ def read(self):
 
         :returns: Nodes per row of the study file
         """
-        builder = _StudyRowBuilder(self.header, self.filename, self.study_id)
+        builder = _StudyRowBuilder(self.header, self._filename, self.study_id)
         while True:
             line = self._read_next_line()
             if line:
@@ -942,7 +942,7 @@ def __init__(self, study_id: str, assay_id: str, input_file: TextIO, filename: O
         self.study_id = study_id
         self.assay_id = assay_id
         self.input_file = input_file
-        self.filename = filename or getattr(input_file, "name", "<no file>")
+        self._filename = filename or getattr(input_file, "name", "<no file>")
         self.unique_rows = set()
         self.duplicate_rows = []
         self._reader = csv.reader(input_file, delimiter="\t", quotechar='"')
@@ -974,7 +974,7 @@ def _read_next_line(self):
         except StopIteration:
             self._line = None
         except UnicodeDecodeError as e:  # pragma: no cover
-            msg = f"Invalid encoding of assay file '{self.input_file.name}' (use Unicode/UTF-8)."
+            msg = f"Invalid encoding of assay file '{self._filename}' (use Unicode/UTF-8)."
             raise ParseIsatabException(msg) from e
         return prev_line
 
@@ -984,7 +984,7 @@ def read(self):
 
         :return: Nodes per row of the assay file
         """
-        builder = _AssayRowBuilder(self.header, self.filename, self.study_id, self.assay_id)
+        builder = _AssayRowBuilder(self.header, self._filename, self.study_id, self.assay_id)
         while True:
             line = self._read_next_line()
             if line: